|
Activity Number:
|
17
- Technology Impact on Total Survey Error
|
|
Type:
|
Invited
|
|
Date/Time:
|
Monday, August 3, 2020 : 10:00 AM to 11:50 AM
|
|
Sponsor:
|
Survey Research Methods Section
|
|
Abstract #308033
|
|
|
Title:
|
Automatic Coding of Open-Ended Questions: Does Double Coding of the Training Data Reduce the Error of Automatic Coding?
|
|
Author(s):
|
Zhoushanyue He and Matthias Schonlau*
|
|
Companies:
|
University of Waterloo and University of Waterloo
|
|
Keywords:
|
statistical learning ;
open-ended questions;
text data;
double-coding
|
|
Abstract:
|
Responses to open-ended questions in surveys are often coded into pre-specified classes, manually or automatically using a statistical learning algorithm. Automatic coding of open-ended responses relies on a set of manually coded responses, based on which a statistical learning model is fitted. Both automatic and manual coding is subject to error. We investigate whether and how manual double coding can reduce error in the automatic classification of open-ended responses. We evaluate several strategies for training the statistical algorithm on double coded data, using experiments on simulated data and real data.
|
Authors who are presenting talks have a * after their name.
